yeah, for the most part i'm reading the code, though, with this test writing not so much... i mean, it's just reading the code itself, establishing all the expected and possible inputs and outputs and writing validation tests that everything is handled. it is very nice because it tends to find places where edge cases could cause quite a big problem. in Go you always aim to fail fast - to handle error cases before you handle success cases. the AI makes it really easy to be extremely comprehensive, it can definitely type faster than me stuff that is usually more or less the same shit over and over again, with different names and mixes of things relevant to every function.

i will be reading the code more as i get to tracing the bugs tho